Exploiting TCF3-Driven Oncogenic Circuits in Burkitt Lymphoma

Burkitt lymphoma (BL) is an aggressive B-cell lymphoma derived from transformed dark zone germinal center B cells. Intensive polychemotherapy in combination with the CD20-directed monoclonal antibody (mAb) rituximab is used as first-line therapy with high cure rates. The prognosis of patients significantly decreases in elderly and unfit patients or with relapsed/refractory disease, underscoring that novel, well-tolerated treatment options are an area of unmet medical need in BL.

Genetically, BLs are characterized by the t(8;14) translocation that juxtaposes the MYC oncogene to the immunoglobulin heavy chain locus (IGH), enhancing the MYC protein expression. Up to 70% of BL also exhibit mutations in TCF3 or its negative regulator, ID3, eventually activating the transcriptional activity of TCF3, also known as E2A. Since both genetic hallmark alterations in BL, MYC-translocation and TCF3 mutations, are currently undruggable, we aimed to identify the E2A-associated oncogenic circuits and exploit potentially druggable downstream targets.

Using chromatin immunoprecipitation sequencing (ChIP-Seq) for E2A in a panel of 10 preclinical BL cell lines, including two newly generated patient-derived BL cell lines (PDBLCL, UPF9T and UPF33P), and 5 diffuse large B-cell lymphoma (DLBCL) control cell lines, 3 germinal center B-cell-like DLBCL and 2 activated B-cell-like DLBCL, we identified the BL-specific E2A oncogenic transcriptional circuits. Pathway enrichment on the top 500 BL-specific ChIP-Seq peaks revealed, among others, frequent binding of E2A to the promoter of multiple components of the B-cell antigen receptor (BCR) signaling cascade. We confirmed the uniform expression of the proximal BCR signaling machinery, including CD79A, CD79B and CD19, and demonstrated functionality using western blot and calcium signaling following crosslinking of the BCR in all cell lines except Raji and UPF9T, which are lacking the BCR and served for subsequent functional experiments as a negative control. Subsequently, we targeted the BCR signaling cascade with inhibitors of proximal BCR signaling, including the PI3Kα/δ inhibitor copanlisib, and demonstrated sensitivity in all cell lines with functional BCR. Using BH3 profiling, we showed that most of the available BL cell lines and the BCR-positive UPF33P PDBLCL were MCL1 dependent and this MCL1 dependency was effectively targeted by PI3Kα/δ inhibition.

Interestingly, we found CD38 among the significant ChIP-Seq peaks, prompting us to follow up on its potential role as a druggable target in BL. First, we confirmed that BL exhibited the highest transcript abundance across 1400 cancer cell lines. We also confirmed that BL exhibited homogenously high expression of CD38 in 10 primary BL samples, compared to a panel of 238 other primary tumor tissues measured on tissue microarrays, including multiple myeloma, follicular lymphoma, DLBCL, mantle cell lymphoma, chronic lymphocytic leukemia, lymphoplasmacytoid lymphoma, marginal zone lymphoma, B-acute lymphoblastic leukemia, Hodgkin lymphoma and primary mediastinal B-cell lymphoma. Next, we confirmed with dual luciferase reporter constructs that CD38 is at least partly driven by hyperactive E2A. Besides E2A binding to the promoter region, we identified two intronic enhancer regions (Enh I and Enh II) that were bound by E2A in BL, underscoring that the binding to the promoter and both enhancer regions is essential for the high CD38 expression in BL. Most interestingly, using three novel patient-derived xenograft models of BL patients in NOD SCID gamma (NSG) mice, we demonstrated that a combination of the CD20- and CD38-directed mAbs, rituximab and daratumumab, significantly decreases tumor growth and increased overall survival of NSG mice compared to mice single agent treated or vehicle control. In addition, we developed a novel antibody-drug conjugate that linked the antineoplastic agent monomethyl auristatin E (MMAE) to daratumumab and demonstrated its applicability in vitro and in vivo.

Our functional genomics-based study revealed proximal inhibitors of BCR signaling and anti-CD38 antibodies as druggable TCF3-driven transcriptional targets in BL warranting further clinical testing. Finally, this study highlights the relevance of the approach to search for druggable downstream targets of undruggable, genetically-driven upstream regulators, which can also be applied to other diseases.